In general, in an active matrix type image display device, an image is displayed by arranging a large number of pixels in a matrix and controlling a light intensity for every pixel in accordance with given brightness information. When using a liquid crystal as an electro-optical substance, the transmittance of each pixel varies in accordance with a voltage written into the pixel. In an active matrix type image display device using an organic electroluminescence (EL) material as the electro-optical substance as well, the basic operation is similar to that of the case where a liquid crystal is used. However, unlike a liquid crystal display, an organic EL display is a so-called self-luminescent type having a light emitting element for every pixel, so has the advantages of a better visual recognition of the mage in comparison with a liquid crystal display, no need for back light, and a fast response speed. The brightnesses of individual light emitting elements are controlled by the amount of current. Namely, this display is largely different from a liquid crystal display in the point that the light emitting elements are current driven types or current controlled types.
In the same way as a liquid crystal display, in an organic EL display as well, there are a simple matrix and an active matrix drive methods. The former is simple in structure, but makes it difficult to realize a large sized, high definition display, so the active matrix method is being vigorously developed. The active matrix method controls the current flowing through the light emitting element provided in each pixel by an active element (generally a thin film transistor, one type of the insulating gate type field effect transistor, hereinafter sometimes referred to as a “TFT”) provided inside the pixel. An organic EL display of this active matrix method is disclosed in for example Japanese Unexamined Patent Publication (Kokai) No. 8-234683. One pixel's worth of an equivalent circuit is shown in FIG. 1. The pixel is comprised of a light emitting element OLED, a first thin film transistor TFT1, a second thin film transistor TFT2, and a holding capacitor C. The light emitting element is an organic electroluminescence (EL) element. An organic EL element has a rectification property in many cases, so is sometimes referred to as an OLED (organic light emitting diode). In the figure, the symbol of a diode is used to indicate the light emitting element OLED. However, the light emitting element is not always limited to an OLED and may be any element controlled in brightness by the amount of the current flowing through it. Also, a rectification property is not always required in the light emitting element. In the illustrated example, a source of the TFT2 is set at a reference potential (ground potential), an anode of the light emitting element OLED is connected to Vdd (power supply potential), and a cathode is connected to a drain of the TFT2. On the other hand, a gate of the TFT1 is connected to a scanning line SCAN, the source is connected to a data line DATA, and the drain is connected to the holding capacitor C and the gate of the TFT2.
In order to operate the pixel, first, when the scanning line SCAN is brought to a selected state and a data potential Vw representing the brightness information is applied to the data line DATA, the TFT1 becomes conductive, the holding capacitor C is charged or discharged, and the gate potential of the TFT2 coincides with the data potential Vw. When the scanning line SCAN is brought to an unselected state, the TFT1 becomes OFF and the TFT2 is electrically separated from the data line DATA, but the gate potential of the TFT2 is stably held by the holding capacitor C. The current flowing through the light emitting element OLED via the TFT2 becomes a value in accordance with a gate/source voltage Vgs, and the light emitting element OLED continuously emits the light with a brightness in accordance with the amount of the current supplied through the TFT2.
When the current flowing between the drain and source of the TFT2 is Ids, this is the drive current flowing through the OLED. Assuming that the TFT2 operates in the saturated region, Ids is represented by the following equation.Ids=μ·Cox·W/L/2(Vgs−Vth)2=μ·Cox·W/L/2(Vw−Vth)22  (1) 
Here, Cox is the gate capacity per unit area and is given by the following equation:Cox=ε0·εr/d  (2) In equation (1) and equation (2), Vth indicates a threshold value of the TFT2, μ indicates a mobility of a carrier, W indicates a channel width, L indicates a channel length, ε0 indicates a permittivity of vacuum, er indicates a dielectric constant of the gate insulating film, and d is a thickness of the gate insulating film.
According to equation (1), Ids can be controlled by the potential Vw written into the pixel. As a result, the brightness of the light emitting element OLED can be controlled. Here, the reason for the operation of the TFT2 in the saturated region is as follows. Namely, this is because, in the saturated region, Ids is controlled by only the Vgs and does not depend upon the drain/source voltage Vds. Therefore, even if Vds fluctuates due to variations in the characteristics of the OLED, a predetermined amount of the drive current Ids can be passed through the OLED.
As mentioned above, in the circuit configuration of the pixel shown in FIG. 1, when written by Vw once, the OLED continues emitting light with a constant brightness during one scanning cycle (one frame) until next rewritten. If large number of such pixels are arranged in a matrix as in FIG. 2, an active matrix type display device can be configured. As shown in FIG. 2, in a conventional display device, scanning lines SCAN-1 through SCAN-N for selecting pixels 25 in a predetermined scanning cycle (for example a frame cycle according to an NTSC standard) and data lines DATA giving brightness information (data potential Vw) for driving the pixels 25 are arranged in a matrix. The scanning lines SCAN-1 through SCAN-N are connected to a scanning line drive circuit 21, while the data lines DATA are connected to a data line drive circuit 22. By repeating the writing of Vw from the data lines DATA by the data line drive circuit 22 while successively selecting the scanning lines SCAN-1 through SCAN-N by the scanning line drive circuit 21, an intended image can be displayed. In a simple matrix type display device, the light emitting element contained in each pixel emits light only at an instant of selection. In contrast, in the active matrix type display device shown in FIG. 2, the light emitting element of each pixel 25 continues to emit light even after finishing being written. Therefore, in particular in a large sized, high definition display, there is the advantage that the level of the drive current of the light emitting elements can be lowered in comparison with the simple matrix type.
FIG. 3 schematically shows a sectional structure of the pixel 25 shown in FIG. 2. Note, only OLED and TFT2 are represented for facilitating the illustration. The OLED is configured by successively superposing a transparent electrode 10, an organic EL layer 11, and a metal electrode 12. The transparent electrode 10 is separated for every pixel, acts as the anode of the OLED, and is made of a transparent conductive film for example ITO. The metal electrode 12 is commonly connected among pixels and acts as the cathode of the OLED. Namely, the metal electrode 12 is commonly connected to a predetermined power supply potential Vdd. The organic EL layer 11 is a composite film obtained by superposing for example a positive hole transport layer and an electron transport layer. For example, Diamyne is vapor deposited on the transparent electrode 10 acting as the anode (positive hole injection electrode) as the positive hole transport layer, Alq3 is vapor deposited thereon as the electron transport layer. Further, a metal electrode 12 acting as the cathode (electron injection electrode) is grown thereon. Note that, Alq3 represents 8-hydroxy quinoline aluminum. The OLED having such a laminate structure is only one example. When a voltage in a forward direction (about 10V) is applied between the anode and the cathode of the OLED having such a configuration, injection of carriers such as electrons and positive holes occurs and luminescence is observed. The operation of the OLED can be considered to be the emission of light by excisions formed by the positive holes injected from the positive hole transport layer and the electrons injected from the electron transport layer.
On the other hand, the TFT2 comprises a gate electrode 2 formed on a substrate 1 made of glass or the like, a gate insulating film 3 superimposed on the top surface thereof, and a semiconductor thin film 4 superimposed above the gate electrode 2 via this gate insulating film 3. This semiconductor thin film 4 is made of for example a polycrystalline silicon thin film. The TFT2 is provided with a source S, a channel Ch, and a drain D acting as a passage of the current supplied to the OLED. The channel Ch is located immediately directly above the gate electrode 2. The TFT2 of this bottom gate structure is coated by an inter-layer insulating film 5. A source electrode 6 and a drain electrode 7 are formed above that. Above them, the OLED mentioned above is grown via another inter-layer insulating film 9. Note that, in the example of FIG. 3, the anode of the OLED is connected to the drain of the TFT2, so a P-channel thin film transistor is used as the TFT2.
In an active matrix type organic EL display, generally a TFT (thin film transistor) formed on a glass substrate is utilized as the active element. This is for the following reason. Namely, an organic EL display is a direct viewing type. Due to this, it becomes relatively large in size. Due to restrictions of cost and manufacturing facilities, a usage of a single crystalline silicon substrate for the formation of the active elements is not practical. Further, in order to extract the light from the light emitting elements, usually a transparent conductive film of ITO (indium tin oxide) is used as the anode of the organic EL layer, but ITO is frequently generally grown under a high temperature which an organic EL layer cannot endure. In this case, it is necessary to form the ITO before the formation of the organic EL layer. Accordingly, the manufacture process roughly becomes as follows:
Referring to FIG. 3 again, first the gate electrode 2, gate insulating film 3, and semiconductor thin film 4 comprised of amorphous silicon are successively stacked and patterned on the glass substrate 1 to form the TFT2. In certain cases, the amorphous silicon is sometimes formed into polysilicon (polycrystalline silicon) by heat treatment such as laser annealing. In this case, generally a TFT2 having a larger degree of carrier mobility in comparison with amorphous silicon and a larger current driving capability can be formed. Next, an ITO transparent electrode 10 acting as the anode of the light emitting element OLED is formed. Subsequently, an organic EL layer 11 is stacked to form the light emitting element OLED. Finally, the metal electrode 12 acting as the cathode of the light emitting element is formed by a metal material (for example aluminum).
In this case, the extraction of the light is started from a back side (bottom surface side) of the substrate 1, so a transparent material (usually a glass) must be used for the substrate 1. In view of this, in an active matrix type organic EL display, a relatively large sized glass substrate 1 is used. As the active element, ordinarily use is made of a TFT as it can be relatively easily formed thereon. Recently, attempts have also been made to extract the light from a front side (top surface side) of the substrate 1. The sectional structure in this case is shown in FIG. 4. The difference of this from FIG. 3 resides in that the light emitting element OLED is comprised by successively superposing a metal electrode 12a, an organic EL layer 11, and a transparent electrode 10a and an N-channel transistor is used as the TFT2.
In this case, the substrate 1 does not have to be transparent like glass, but as the transistor formed on a large sized substrate, use is generally still made of a TFT. However, the amorphous silicon and polysilicon used for the formation of the TFT have a worse crystallinity in comparison with single crystalline silicon and have a poor controllability of the conduction mechanism, therefore it has been known that there is a large variation in characteristics in formed TFTs. Particularly, when a polysilicon TFT is formed on a relatively large sized glass substrate, usually the laser annealing method is used as mentioned-above in order to avoid the problem of thermal deformation of the glass substrate, but it is difficult to uniformly irradiate laser energy to a large glass substrate. Occurrence of variations in the state of the crystallization of the polysilicon according to the location in the substrate cannot be avoided.
As a result, it is not rare for the Vth (threshold value) to vary according to pixel by several hundreds of mV, in certain cases, 1V or more, even in the TFTs formed on an identical substrate. In this case, even if a same signal potential Vw is written with respect to for example different pixels, the Vth will vary according to the pixels. As a result, according to equation (1) described above, the current Ids flowing through the OLEDs will largely vary for every pixel and consequently become completely off from the intended value, so a high quality of image cannot be expected as the display. A similar thing can be said for not only the Vth, but also the variation of parameters of equation (1) such as the carrier mobility A. Further, a certain degree of fluctuation in the above parameters is unavoidable not only due to the variation among pixels as mentioned above, but also variations for every manufacturing lot or every product. In such a case, it is necessary to determine how the data line potential Vw should be set with respect to the intended current Ids to be passed through the OLEDs for every product in accordance with the final state of the parameters of equation (1). Not only is this impractical in the mass production process of displays, but it is also extremely difficult to devise countermeasures for fluctuations in characteristics of the TFTs due to the ambient temperature and changes of the TFT characteristics occurring due to usage over a long period of time.